One-Way Anova Application, Essay Example

The article of Burke discusses the propriety of using different kinds of ANOVA tests as compared to t-tests in cases when analysis of variance of one and several indicators has to be assessed. The author states that ANOVA is successful when applied to analysis of several data sets simultaneously for the sake of achieving a homogeneous decision on the measure of variance among them (Burke 9). ANOVA helps answer the question on whether there is significant difference between mean levels under the condition that these values have been received by means of replicate observations. “‘Significant’ refers to the observed spread of means that would not normally arise from the chance variation within groups” (Burke 8).

The one-way ANOVA can be used in cases when only one factor is considered and replicate data from changing the level is available. The indisputable advantage of one-way ANOVA for effective decision-making on the basis of given statistical data is that the number of measurements is much lower than for conducting t-tests (Burke 9). However, this fact is also one of the major pitfalls of ANOVA as lack of numerous indicators does not allow the analyst to identify the measure of departure from normality (Burke 9).

Some suggested applications for one-way ANOVA not discussed in the present work of Burke include various situations with one factor at several levels and multiple observations made for these several levels. The one-way ANOVA in this case can allow analysis of variation both within levels and across levels, thus providing a much more precise decision-making tool for analysts. As in case of assessment of task completion time depending on the employee level of preparation: the single factor assessed is the time of task completion, but the levels chosen for the indicator may include the beginner, the intermediate and professional level. Each level has a sample of employees chosen for measurement, thus providing multiple measurements for each level. Finally, these three independent groups are assessed for variance within and across to achieve two conclusions at once – whether the variance within the group affects the overall result and whether the variance across groups supports the hypothesis about the professional group accomplishing the task quicker.

Works Cited

Burke, Shaun. ‘Analysis of Variance’. LC•GC Europe Online Supplement. LGC, Teddington, Middlesex, UK.